Computer vs Board (was Re: PBM design and formulae - a fresh start)

From: troyer@cgl.ucsf.edu (John M. Troyer)

Date: Tue, 06 Jul 1993 18:14:46 +0000

>desj@ccr-p.ida.org (David desJardins) wrote:
>>Computers don't give a nice 'reality' feeling unless you are completely
>>blind to what is going on behind the screen.  The best strategy
>>boardgames (like ASL) are much more realistic than the best computer
>>strategy games, just because they are more complex and can model more
>>stuff.  Computer games are just not sophisticated enough to be realistic
>>in the way you want them to be.  Who knows, that might happen one day.
>>But it is a long long way off.

bc@lnec.pt (Luis Miguel Sequeira) writes:
>Which means  that  a game  designed to be  played by humans to act  as
>referees  (eg. ASL) is too complicated for our  poor average computers
>of the '90s?

It means that humans have this great exception-handling, data-scoping, and
sanity-checking ability that lets them apply the correct rules at the
correct time based on the intent of the game designer. To set down the
rules in code & make sure they are self-consistent and their interactions
don't introduce nonsensical outcomes is difficult. Not to mention
errors in the implementation (ie bugs).

I work with 100,000 lines of crufty 1970's fortran code to do simulations.
Bugs are still emerging. And you expect some ASL code to work out of the
box? The rules probably aren't even all self-consistent (well, maybe
for ASL, the years of wargamers have probably fixed most of those.)
Somebody mentioned Hero was rewriting their rules a bit as a result of making
a computerized combat simulator.

What David says stands. No CURRENTLY IMPLEMENTED computer games come
close. Mostly because it's not economically feasible. It takes too
long to program/debug.

sweeping generalizations:

Board-wargames are marketed for anal retentive wargamers. They may have
complicated rules, but they are extensively playtested for bad rules.
(Bad rules interfere with play/are illogical/destroy game balance/etc.)

Commercial mac/pc computer games are marketed for 12 year olds. They try
not to be very complex, and are sometimes playtested for both bad rules and
bugs.  (sometimes not, eg Pax Imperia)

Non-commercial computer games are written by students or bored programmers.
They tend to be complex (eg empire, galaxy, gb, nethack, et al.), since
coding is easier than playtesting. Since they cannot swipe playtested
boardgame rules, they have a tendancy to have bad rules. They always have bugs.

Scientific simulation packages are written by lots and lots of grad students
and postdocs over the years (in contrast to the above, which are written
by only a few people.) They are bug-tested by hundreds of scientists. They
use up millions in grants. They still have bugs.

Now, when you get a few million in grants to code up an ASL simulation, give
me a call. I'll pay $49.95 for a copy.

john

ps I know I'm comparing apples & oranges here. Don't nitpick. It says
'sweeping generalizations', ok? But I do think that something more complex
than any boardgame could be coded up. And the complexity hidden so that it
was playable. I just don't think it's economically feasible to do so right
now.

pps Luis, if you don't quit right-justifying your text, I'm going to scream.

References

Re: PBM design and formulae - a fresh start
bc@lnec.pt (Luis Miguel Sequeira)
Tue, 06 Jul 1993 17:49:51 +0000